This poster will present the preliminary results of my PhD project which researches trends in the early Gothic novel using computational methods. A corpus of over 2500 British early Gothic texts has been created and explored using a combination of the analysis of word embeddings and sub-sections of the corpus defined by annotated meta-data. Variations in the embeddings for such sub-corpora demonstrate that various established theories such as the assumed British preoccupation with European national identity in Gothic fiction, the tendency to categorise all Gothic novels as filled with negative sentiment, or the ambiguously defined 'female Gothic', warrant re-evaluation and further exploration assisted by quantitative methods. Notable changes in embeddings for specific datasets such those comprised of texts written at the onset of the Gothic's popularity, or by female authors, are visualised in this poster.